XPK Start: Sat Apr 18 13:03:14 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-18 13:03:53.239943: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303)
I0418 13:03:53.420944 136958621275968 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 13:04:02.465175 136958621275968 distributed.py:172] Connecting to JAX distributed service on mt-01-base-n1jit-slice-job-0-0.mt-01-base-n1jit:8482
I0418 13:04:07.766326 136958621275968 max_utils.py:284] Jax distributed system initialized!
I0418 13:04:14.521549 136958621275968 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 13:04:14.521653 136958621275968 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 13:04:14.521694 136958621275968 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0418 13:04:14.521729 136958621275968 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 13:04:15.221219 136958621275968 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 13:04:15.221511 136958621275968 checkpointing.py:677] Setting up checkpoint logger...
I0418 13:04:15.221564 136958621275968 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 13:04:15.221607 136958621275968 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 13:04:15.221955 136958621275968 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7c8f933b1520>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 13:04:19.681347 136958621275968 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 13:04:19.681578 136958621275968 checkpoint_manager.py:708] [process=2][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c7cb8103230>}, handler_registry=None
I0418 13:04:19.681815 136958621275968 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c7cb8103230>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`.
I0418 13:04:19.681862 136958621275968 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c7cb8104800>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`.
I0418 13:04:19.681898 136958621275968 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c7cb8103230>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c7cb8103230>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c7cb8104800>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c7cb8104800>}).
I0418 13:04:19.682215 136958621275968 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 13:04:19.682296 136958621275968 async_checkpointer.py:192] [process=2][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c7c985a00e0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 13:04:21.742517 136958621275968 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints
I0418 13:04:23.045755 136958621275968 checkpoint_manager.py:929] [process=2][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c7cb8106630>
I0418 13:04:23.045926 136958621275968 checkpointing.py:301] Checkpoint manager created!
I0418 13:04:23.971249 136958621275968 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 13:04:23.971367 136958621275968 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:04:24.348458 136958621275968 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:04:24.348549 136958621275968 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:04:24.364951 136958621275968 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:04:24.365013 136958621275968 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:04:24.388924 136958621275968 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:04:24.388990 136958621275968 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:04:24.405544 136958621275968 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:04:24.405606 136958621275968 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:04:24.421940 136958621275968 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:04:24.421999 136958621275968 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:04:24.446753 136958621275968 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 13:04:24.446823 136958621275968 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:04:24.468928 136958621275968 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 13:04:24.468993 136958621275968 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 13:04:24.688409 136958621275968 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 13:04:24.688515 136958621275968 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0418 13:04:26.134182 136958621275968 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 13:04:26.134322 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 13:04:26.134377 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 13:04:26.134434 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0418 13:04:26.134487 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 13:04:26.134530 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 13:04:26.134609 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 13:04:26.134683 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0418 13:04:26.134728 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0418 13:04:26.134765 136958621275968 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 13:04:26.134821 136958621275968 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 13:04:26.134871 136958621275968 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 13:04:26.630404 136958621275968 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:04:26.630499 136958621275968 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 13:04:26.646153 136958621275968 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:04:26.646215 136958621275968 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 13:04:37.026200 136958621275968 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0418 13:04:37.027051 136958621275968 metric_logger.py:301] number parameters: 1.104 billion
I0418 13:04:47.908642 136958621275968 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0418 13:04:49.359531 136958621275968 checkpointing.py:776] Waited 1.4508686065673828 seconds for step 0 to finish before starting checkpointing.
I0418 13:04:49.362021 136958621275968 checkpoint_manager.py:2009] [process=2][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:04:49.364090 136958621275968 checkpoint_manager.py:1512] [process=2] Saving checkpoint at step 0
I0418 13:04:49.365388 136958621275968 event_tracking.py:70] [process=2] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints/0.
I0418 13:04:50.121481 136958621275968 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 13:04:50.122584 136958621275968 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 13:04:50.122641 136958621275968 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0418 13:04:50.394829 136958621275968 base_pytree_checkpoint_handler.py:154] [process=2][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.273408s
I0418 13:04:50.395011 136958621275968 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/blocking_gbytes_per_sec: 5.532 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2788264751434326 s) (per-host)
I0418 13:04:50.395066 136958621275968 base_pytree_checkpoint_handler.py:768] [process=2][thread=MainThread] Initiated Pytree async_save. Time taken: 0.278896s (batch_requests_ready=0.002241s, total_serialization_initiated=0.276579s, others=0.000076s)
I0418 13:04:50.395160 136958621275968 composite_checkpoint_handler.py:715] [process=2][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.282783s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=0.282765)
I0418 13:04:50.395978 136958621275968 event_tracking.py:125] [process=2] [async] Finished blocking save in 1.03 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints/0.
I0418 13:04:50.396327 136815147214592 async_checkpointer.py:76] [process=2][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:24:50.396282
I0418 13:04:50.405106 136958621275968 checkpoint_manager.py:1560] [process=2][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:04:50.405378 136819529488128 async_checkpointer.py:280] [process=2][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:04:50.405505 136958621275968 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776517489.362002, 'wait_for_prev_duration_secs': 6.198883056640625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776517489.3641284, 'checkpointer_blocking_duration_secs': 1.0323567390441895, 'get_old_steps_start_time': 1776517490.3965104, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776517489.3600717, 'checkpoint_manager_blocking_duration_secs': 1.045393466949463}
I0418 13:04:50.405674 136958621275968 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0418 13:04:50.405727 136958621275968 max_utils.py:750]
Memstats: After params initialized:
I0418 13:04:50.405778 136958621275968 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_8(process=2,(0,2,0,0))
I0418 13:04:50.405808 136958621275968 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_9(process=2,(1,2,0,0))
I0418 13:04:50.405834 136958621275968 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_12(process=2,(0,3,0,0))
I0418 13:04:50.405856 136958621275968 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_13(process=2,(1,3,0,0))
I0418 13:04:50.721762 136958621275968 metric_logger.py:196] completed step: 0, seconds: 10.881, TFLOP/s/device: 1.249, Tokens/s/device: 188.210, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0418 13:04:50.891072 136958621275968 metric_logger.py:196] completed step: 1, seconds: 2.812, TFLOP/s/device: 4.833, Tokens/s/device: 728.435, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0418 13:04:51.298757 136958621275968 metric_logger.py:196] completed step: 2, seconds: 0.022, TFLOP/s/device: 618.159, Tokens/s/device: 93175.614, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0418 13:04:51.446011 136958621275968 metric_logger.py:196] completed step: 3, seconds: 0.408, TFLOP/s/device: 33.310, Tokens/s/device: 5020.814, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0418 13:04:51.741594 136958621275968 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.539, Tokens/s/device: 13345.497, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0418 13:04:51.747609 136958621275968 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.290, Tokens/s/device: 13910.964, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0418 13:04:56.813367 2570 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 13:05:01.511497 136816528623360 array_metadata_store.py:203] [process=2][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints/0/items/array_metadatas/process_2
I0418 13:05:16.296361 136958621275968 metric_logger.py:196] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.927, Tokens/s/device: 6922.684, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0418 13:05:16.443649 136958621275968 metric_logger.py:196] completed step: 7, seconds: 24.402, TFLOP/s/device: 0.557, Tokens/s/device: 83.928, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0418 13:05:16.591010 136958621275968 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 89.057, Tokens/s/device: 13423.699, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0418 13:05:16.737620 136958621275968 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0418 13:05:16.738307 136958621275968 checkpointing.py:776] Waited 0.0007073879241943359 seconds for step 9 to finish before starting checkpointing.
I0418 13:05:16.741471 136958621275968 checkpoint_manager.py:2020] [process=2][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:05:29.757964 136815147214592 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/gbytes_per_sec: 39.847 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 39.64174437522888 s) (per-host)
I0418 13:05:29.758090 136815147214592 async_checkpointer.py:90] [process=2][thread=async_save] 3 Handler Commit operations completed. Time taken: 39.361650s.
I0418 13:05:39.683833 136815147214592 async_checkpointer.py:160] [process=2][thread=async_save] Background save thread done. Time taken: 49.287375s.
I0418 13:05:39.684109 136819529488128 async_checkpointer.py:288] [process=2][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:05:39.684228 136819529488128 async_checkpointer.py:298] [process=2][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:05:39.684308 136819529488128 checkpoint_manager.py:2137] [process=2][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:05:39.686467 136819529488128 checkpoint_manager.py:2146] [process=2][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0418 13:05:39.686657 136958621275968 checkpoint_manager.py:2032] [process=2][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0418 13:05:39.686790 136958621275968 checkpoint_manager.py:1452] Waiting for previous save to complete took 22.945321 seconds. If this number is high, consider checkpointing less frequently.
I0418 13:05:39.688615 136958621275968 checkpoint_manager.py:1512] [process=2] Saving checkpoint at step 9
I0418 13:05:39.690660 136958621275968 event_tracking.py:70] [process=2] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints/9.
I0418 13:05:40.843220 136958621275968 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 13:05:40.843329 136958621275968 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0418 13:05:40.874107 136958621275968 base_pytree_checkpoint_handler.py:154] [process=2][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.032105s
I0418 13:05:40.874300 136958621275968 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/blocking_gbytes_per_sec: 43.411 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03553438186645508 s) (per-host)
I0418 13:05:40.874360 136958621275968 base_pytree_checkpoint_handler.py:768] [process=2][thread=MainThread] Initiated Pytree async_save. Time taken: 0.035605s (batch_requests_ready=0.001774s, total_serialization_initiated=0.033751s, others=0.000079s)
I0418 13:05:40.874454 136958621275968 composite_checkpoint_handler.py:715] [process=2][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.039627s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.039612)
I0418 13:05:40.875200 136958621275968 event_tracking.py:125] [process=2] [async] Finished blocking save in 1.19 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints/9.
I0418 13:05:40.875556 136819529488128 async_checkpointer.py:76] [process=2][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:25:40.875520
I0418 13:05:40.879975 136958621275968 checkpoint_manager.py:1560] [process=2][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:05:40.880220 136816528623360 async_checkpointer.py:280] [process=2][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:05:40.880364 136958621275968 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776517516.7414389, 'wait_for_prev_duration_secs': 22.945321083068848, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776517539.688655, 'checkpointer_blocking_duration_secs': 1.1870481967926025, 'get_old_steps_start_time': 1776517540.8757267, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776517516.7385318, 'checkpoint_manager_blocking_duration_secs': 24.141796350479126}
I0418 13:05:40.880470 136958621275968 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0418 13:05:40.880516 136958621275968 checkpoint_manager.py:2020] [process=2][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:05:50.775071 136805361866496 array_metadata_store.py:203] [process=2][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_01_base/checkpoints/9/items/array_metadatas/process_2
I0418 13:06:24.945389 136819529488128 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/gbytes_per_sec: 35.813 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.10658097267151 s) (per-host)
I0418 13:06:24.945522 136819529488128 async_checkpointer.py:90] [process=2][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.069849s.
I0418 13:06:32.245451 136819529488128 async_checkpointer.py:160] [process=2][thread=async_save] Background save thread done. Time taken: 51.369763s.
I0418 13:06:32.245718 136816528623360 async_checkpointer.py:288] [process=2][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:06:32.245847 136816528623360 async_checkpointer.py:298] [process=2][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:06:32.245898 136816528623360 checkpoint_manager.py:2137] [process=2][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:06:32.247511 136816528623360 checkpoint_manager.py:2146] [process=2][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0418 13:06:32.247698 136958621275968 checkpoint_manager.py:2032] [process=2][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0418 13:06:32.247846 136958621275968 checkpoint_manager.py:2009] [process=2][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:06:32.248755 136958621275968 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.337, Tokens/s/device: 13918.055, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 13:06:45 UTC 2026
EXIT_CODE=0